Magnesium transition-metal based lightweight hydrogen
storage materials : a NMR study
Citation for published version (APA):
Srinivasan, S. (2011). Magnesium transition-metal based lightweight hydrogen storage materials : a NMR study. Technische Universiteit Eindhoven. https://doi.org/10.6100/IR710917
DOI:
10.6100/IR710917
Document status and date: Published: 01/01/2011 Document Version:
Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:
www.tue.nl/taverne Take down policy
If you believe that this document breaches copyright please contact us at: openaccess@tue.nl
Magnesium Transition-Metal based lightweight hydrogen-storage materials. A
NMR study
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de
Technische Universiteit Eindhoven, op gezag van de
rector magnificus, prof.dr.ir. C.J. van Duijn, voor een
commissie aangewezen door het College voor
Promoties in het openbaar te verdedigen
op maandag 14 maart 2011 om 16.00 uur
door
Subramanian Srinivasan
II
Dit proefschrift is goedgekeurd door de promotoren:
prof.dr. R.A. van Santen
en
prof.dr. P.H.L. Notten
Copromotor:
III
Chapter 1: Introduction and Scope 1.1 Introduction
1.2 Hydrogen economy 1.3 Hydrogen storage
1.4 Recent progress in hydrogen storage MgH2
1.5 NMR methodology
1.6 Goal and scope of this thesis 1.7 References
Chapter 2: Characterization techniques 2.1 Introduction
2.2 Nuclear Magnetic Resonance 2.2.1 Spectral editing
2.2.1.1 Magic Angle Spinning 2.2.1.2 Deuteration effect 2.2.1.3 TRAPDOR NMR
2.2.1.4 Deuterium Double Quantum NMR 2.2.2 Mobility from NMR
2.2.2.1 Two- Dimensional Exchange NMR spectroscopy (2D Exsy) 2.2.2.2 One Dimensional Exchange Spectroscopy (1D Exsy) 2.2.2.3 Relaxometry techniques
2.3 Powder X-ray diffraction 2.4 References
Chapter 3: Nanostructures of Mg0.65Ti0.35Dxstudied with X-ray diffraction, neutron diffraction, and magic-angle-spinning 2H NMR spectroscopy
3.1 Introduction
3.2 Experimental section 3.3 Results
3.3.1 XRD characterization
IV
3.5 Conclusions 3.6 References
Chapter 4: Siting and mobility of deuterium absorbed in co-sputtered Mg0.65Ti0.35. A
MAS 2H NMR study 4.1 Introduction 4.2 Experimental 4.2.1 Material preparation 4.2.2 NMR Characterization 4.2.3 Data Analysis
4.3 Results and Discussion
4.3.1 MAS 2H NMR comparison of different metal deuterides
4.3.2 Deuterium dynamics at different timescales and varied temperatures 4.3.3 Mg-Ti nano-segregation model consistent with the 2H NMR results 4.4 Conclusions
4.5 References
Chapter 5: Dark deuterium in lightweight hydrogen-storage material Mg0.65Sc0.35D2.2
studied with MAS 2H NMR 5.1 Introduction
5.2 Theory
5.2.1 Dynamic-equilibrium deuteron exchange between two states with different NMR visibility
5.3 Experimental
5.3.1 Material synthesis 5.3.2 MAS 2H NMR
5.3.3 Double-quantum MAS 2H NMR with 45Sc recoupling 5.3.4 2H-{45Sc} TRAPDOR
5.3.5 2H MAS NMR Exchange Spectroscopy 5.3.6 Lineshape analysis
V
5.4.2 Deuterium exchange at room temperature from 1D and 2D Exsy 5.4.3 Overall peak-area analysis
5.5 Conclusions 5.6 References
Chapter 6: 1H NMR quantification of different hydrogen-containing phases in nano-composites of MgH2 and carbon.
6.1 Introduction 6.2 Experimental
6.3 Results and discussion
6.3.1.1 XRD characterization 6.3.1.2 XRD quantification
6.3.2 Temperature Programmed desorption 6.3.3 NMR characterization
6.3.3.1 Susceptibility and radio frequency penetration effects 6.3.3.2 MAS NMR spectra
6.3.3.3 Static NMR spectra 6.3.3.4 NMR quantification 6.3.3.5 Combined analysis
6.3.3.6 Effect and extent of oxidation
6.3.3.7 Two-dimensional exchange spectroscopy (2D Exsy) 6.4 Conclusion 6.5 References Chapter 7 Summary Appendix Acknowledgements Curriculum Vitae List of Publications
Chapter 1
1.1
Introduction
Since the industrial revolution, the energy demand has increased dramatically together with the word population from ~ 5×1012 kWh in the year 1800 to ~120×1012 kWh in 2000.1 Fossil fuels have been used for transportation, power generation and industrial processes. The finite availability of these is a major concern. Moreover, oil reserves are concentrated in particular places in the world whereas the demand for oil is all over the world. The disadvantage of using fossil fuels is the environmental pollution, especially emission of greenhouse gases. These gases trap heat and are the cause of global warming. This has large negative effects such as thinning of glaciers in polar regions and sea level increase.
1.2 Hydrogen Economy
In a future hydrogen-based economy2,3,4,5 conventional fossil fuels will be replaced by hydrogen. Direct hydrogen combustion like in fuel cells, only produces water and also the output from battery-powered cars do not pollute the environment unlike combustion of fossil fuels. This fact makes the shift towards hydrogen-based economy more favorable. Already, cars based on liquid hydrogen as fuels (BMW) and battery-powered hybrid cars (Toyota) are available in the market. Hydrogen is not a fuel but an energy carrier. Although the amount of pure hydrogen gas in atmosphere is negligible, it is abundantly available in other chemical forms, such as water.
1.3 Hydrogen Storage
The main factors for a material to be considered a potential candidate for hydrogen storage are
High gravimetric and volumetric storage capcacity High recyclability
Fast (de)sorption kinetics Low desorption temperature
Hydrogen can be stored in the form of gas at high pressures in metal cylinders. Another way is to store it in a cryogenic liquid form at, e.g., 20 K, which is the boiling point of liquid hydrogen at 1 bar. The advantage of storing as a liquid is that of hydrogen can be stored at high volumetric density compared to compressed gas. However, the energy required for liquefaction and the boil-off hydrogen loss have also to be considered. These storage methods do not (yet) meet the targets for hydrogen storage set by the US Department of Energy (DOE).6 Hydrogen can also be stored by physisorption in carbon nanotubes7, 8, 9, 10 and metal organic frameworks (MOF).11, 12, 13
An alternative for physical storage as high-pressurized gas or cryogenic liquid is to store hydrogen chemically in metal hydrides or other chemical hydrides. Since hydrogen is stored as atoms rather than molecules, chemical storage can yield higher volumetric capacity. The thermodynamic properties of a metal hydrogen system are obtained from the calorimetric or pressure-composition isotherm (PCI). In a PCI measurement, the Van ’t Hoff plot is constructed to obtain thermodynamic parameters. At equilibrium, the hydrogen pressure and temperature are related by the Van’t Hoff equation, according to
R S RT H P T P eq eq ∆ − ∆ = 0 ) ( ln (1.3.1)
where H∆ , S∆ , R are the formation enthalpy, the entropy change and the universal gas constant, respectively. Ideally, to have an equilibrium pressure of 1 bar at room temperature, the optimum formation enthalpy can be calculated to be ~ – 40 kJ/mol H2. Fig. 1.1 schematically shows the
Pressure-composition isotherms of a typical metal hydride, such as LaNi5. When hydrogen
absorbs into a metal, it first forms a hydrogen-poor solid-solution phase, the so-called α phase. The equilibrium pressure increases as more hydrogen is absorbed in this phase until a critical
concentration (xα) is reached, after which an additional hydrogen-rich phase, called the β phase is
formed. The gas-equilibrium pressure in the two-phase coexistence region remains constant until a second critical concentration (xβ) is reached, where further hydrogen absorption results in the
formation of only β phase and pressure also increases. The coexistence of α and β phases is characterized by plateau. The equilibrium plateau pressure depends on the temperature according to the Van ‘t Hoff equation. (Eq. 1.3.1)
Interstitial hydrides are studied for decades for their reversible hydrogen storage properties. One of the first commercial materials were LaNi5 and FeTi alloys.14 , 15 , 16 These
compounds are formed by combining an element which has a high affinity towards hydrogen with another element that has low affinity towards hydrogen. LaNi5 has an equilibrium pressure
of 1.7 bar at room temperature and it forms LaNi5H6 after complete hydrogenation. In general,
these materials are AB5 type materials, where A is lanthanide element (atomic number 57 – 71)
and B element is Ni, Co, Al, Fe, Sn, Cu, Ti, etc. These materials have superior (de)sorption kinetics but suffer from low gravimetric capacity. They are not suitable for on-board hydrogen storage systems.
Figure 1.1 Pressure-composition isotherms of metal hydrides. (reproduced from Ref
21) The solid solution (α-phase), the hydride phase (β-phase) and the coexistence of the two phases are shown. The coexistence region is reflected by the plateau and exists only below the critical temperature Tc. The construction of the Van't Hoff plot
is shown on the right-hand side. The slope of the line reflects the hydride-formation enthalpy and the intercept the formation entropy . (see Eq. (1.3.1))
Another class of materials is that of the complex hydrides.17,18,19,20 In these hydrides, hydrogen atoms are covalently bonded to a central atom in an anion complex (e.g. [AlH4]-, [BH4]-, [NH2]-)
and are stabilized by a cation, typically an alkaline or alkali earth metal. Depending on the cation and anion complex, the gravimetric storage capacity varies between ~ 6 to 18 wt.% of hydrogen. Fig. 1.2 summarizes the volumetric and gravimetric capacities of hydrogen storage materials. The various classes of hydrogen storage materials along with their desorption temperature are depicted with distinct colors. Each color represents a particular type of hydrogen storage material. High-pressure storage of hydrogen gas has lower volumetric and gravimetric capacities than other methods of hydrogen storage. Moreover, the gravimetric hydrogen density decreases with increasing pressure due to the increasing thickness of the walls of the pressure cylinder.
Hydrogen chemisorbed on carbon, such as in alkanes, in general have a higher gravimetric capacity. However, they produce carbon-dioxide on complete combustion which is a green house gas. Moreover, on incomplete combustion, they also produce carbon-monoxide which is toxic. The gravimetric storage capacity of complex hydrides is greater than in metal hydrides. However, full dehydrogenation takes place in two steps in most of the materials and in some materials reversible storage of hydrogen is an issue. The decomposition temperature is also high. Metal hydrides can reversibly store hydrogen. The most common is the LaNi5 which has good
thermodynamic and kinetic properties. The equilibrium pressure is 2 bars at 300 K. However, it suffers from low gravimetric capacity. In general, metal hydrides with higher gravimetric capacities suffer from thermodynamic and/or kinetic properties.
1.4 Recent progress in hydrogen storage in MgH
2In search for an ideal hydrogen-storage metal hydride, many studies have been done on MgH2. Mg is available in abundance. Its hydride, MgH2, has a high gravimetric capacity and can
store 7.6 wt.% of hydrogen. This property makes MgH2 a potential candidate for hydrogen
storage. However, major challenges have to be overcome in order to put MgH2 into practical use.
MgH2 is thermodynamically highly stable. Its enthalpy of formation is -74 kJ/ mol H2. This
Figure 1.2 Volumetric and gravimetric capacities of different hydrogen storage
materials (reproduced from Ref 21 ). Different colors indicate types of storage materials. Red and pink colors represent hydrogen storage in metal hydrides and complex hydrides. The gravimetric hydrogen storage capacity of complex hydrides is higher than that of the metal hydrides.
K. The goal is to destabilize MgH2 in order to achieve the same equilibrium pressure of 1 bar at
or near room temperature conditions. Apart from high stability, MgH2 also suffers from poor
sorption kinetics. Hydrogen diffusion within MgH2 lattice is slow. Moreover, once the outer
layer of MgH2 is formed, it blocks further hydrogenation of inside layers. Many studies are in
progress to improve kinetic and thermodynamic limitations with different strategies. MgH2 has a
rutile crystal structure with lattice constants a = 4.50 Å and c = 3.01 Å.22 The Wyckoff position of Mg and H are 2a (0, 0, 0) and 4f (0.306, 0.306, 0), respectively.
Notten and his group have improved the kinetics of MgH2 by doping Mg with a
Transition Metal (TM) element.23,24,25,26,27 The addition of minimally 20 at.% of a TM causes a change from the rutile structure of MgH2 to a fluorite structure after hydrogenation. In the
fluorite structure, the hydrogen mobility is higher. Fig. 1.4.2 shows the structure and lattice parameters of rutile and fluorite MgH2 and Mg-TM-H2, respectively. The lattice parameter in the
fluorite structure of Mgx-TM(1-x)H2 depends on the doping percentage of the TM and follows
Vegards law. 28
(a) (b)
Figure 1.4.2. Crystal structures of (a) MgH2 (b) Mg1-xTMxH2. MgH2 crystallizes into
the rutile structure. Green and yellow spheres denote Mg and hydrogen atoms. Upon addition of minimally 20 at.% of a transition element, the crystal structure turns into a fluorite structure after hydrogenation. In the fluorite structure hydrogen atoms located in tetrahedral and octahedral sites represented by yellow and grey spheres, respectively. The red spheres denote either Mg or a TM element.
Starting from MgH2 and TM-H2, high pressure up to 8 GPa and high temperature
(600 °C) conditions were used to obtain Mg6-7TMH14-16 (TM = Ti, Zr, Hf, V, Nb and Ta).29,30,31,32
The resulting material crystal structures are fluorite with lattice constants ~ 4.8 Å. The TM atoms are arranged in an ordered way and can best be described by a superlattice of type Ca7Ge
structure.33 Powder X-ray diffraction using synchrotron radiation revealed two types of tetrahedral sites for hydrogen atoms. At one of the sites, hydrogen is coordinated by three Mg atoms and one TM atom, and in the other site, it coordinates four Mg atoms. These ternary hydrides were initially thought to be metastable as they tend to phase separate into constituent binary hydrides upon complete dehydrogenation.29 Later, these materials were observed to be thermodynamically stable and reversible if not fully dehydrogenated.29
In another approach, nano-structuring is employed to improve the hydrogenation properties. The crystallite size of bulk MgH2 is reduced from micrometer (µm) to nanometer size
(nm) by ball-milling. The kinetics is faster for nano-crystalline MgH2 than for µm-sized
crystallites.34,35 The ball-milling process reduces the crystallite size and as a result the surface area is enhanced. The diffusion path lengths are also decreased as a result of ball-milling. This allows hydrogen atoms to penetrate more easily to form a hydride. The defects and distortions introduced by ball-milling were also considered to be a cause for improved kinetics.35 However, later it was shown that they do not play a decisive role in contributing towards the increased kinetics.36,37 MgH2 ball-milled with TM elements also increased the kinetics but thermodynamic
properties of MgH2 were not affected.38 It was found that the kinetics of ball-milled MgH2 with
Ti or V was higher when compared to ball-milled MgH2 with any other TM elements. MgH2
ball-milled with TM oxides were also found to be very effective in increasing the kinetics of hydrogen (ab) desorption.39,40
Even though crystallite-size reduction enhances the sorption kinetics of MgH2,
agglomeration of particles occurs on subsequent cycles of hydrogen absorption and desorption. This again led to slower (de)sorption kinetics which is comparable to bulk MgH2. A recent
quantum chemical study41 showed that the thermodynamic stability can be altered by reduction of crystallite-size down to ~ 1.3 nm which consisted of at the most 20 metal atoms. De Jongh et.al.42 showed that Mg can be held in nm-sized form in a carbon support. The process is to melt and infiltrate Mg into porous carbon.
In an alternative approach to stabilize the particle size, addition of compounds such as TiF3 was proposed.43,44 The added compound forms a new compound/alloy with MgH2. On
hydrogenation, these alloys act as nucleation centers for the metal hydride phase to be formed. The phase/compound that acts as nucleation centers is known as grain refiner.
Apart from hydrogen storage, Mg-RE-based materials are also studied for technological applications such as hydrogen sensors,45 switchable mirrors or solar collectors.46 Magnetron sputtered Mg-TM films are investigated for the same type of applications.47,48,49,50 Using recently developed Hydrogenography method, the authors proposed a model for MgyTi(1-y) films, wherein,
Mg and Ti are not phase separated, instead a coherent crystal consisting of MgH2 and TiH2
1.5 NMR methodology
In most studies of materials for hydrogen storage, diffraction-based techniques are used to determine the structure, (macroscopic) hydrogen-absorption measurements to determine the thermodynamics properties . These methods give information about the energy carriers in an indirect way. NMR probes the location, as well as the mobility of energy carriers directly. Every characterization technique has its limitations. For nanometer-structured materials with short structure-coherence lengths, XRD yields patterns consisting of broad and overlapping peaks. In contrast, 1H NMR can yield detailed information about hydrogen in XRD-amorphous materials. However, it does not give direct information about the non hydride parts of hydrogen storage materials.
Metal hydrides have been studied with NMR for decades. Especially, diffusion-studies on hydrogen atoms in metal hydrides received much interest.51 , 52 , 53 ,27 Hydrogen mobility at a nanosecond or millisecond timescale was studied with wideline NMR relaxometry techniques. Activation barriers for hydrogen motion were also obtained from temperature variation measurements. Apart from static relaxometry based techniques, Pulsed Field Gradient (PFG) NMR experiments54,55 gave information on diffusion coefficients of hydrogen mobility.
Although the above-mentioned NMR methods were useful to obtain insight into hydrogen mobility, a major problem associated with static 1H NMR of metal hydrides is its broad lines in the spectra. In metal hydrides, there are very short 1H-1H distances. Dipole-dipole interactions among the 1H spins cause strong linebroadening their resonance. Magic-angle spinning (MAS) narrows the resonance. The availability of ultra-fast MAS up to 70 kHz, advanced pulse sequences to obtain resolved lines in the NMR spectra and better spectrometers offer opportunities to view metal hydrides in a different perspective. However, other challenges have to be addressed that concerns NMR on metal hydrides. Eddy currents can be a problem with MAS at high spinning rates. This will force to work at lower MAS rates and hence resolution of spectral lines will be an issue. Moreover, for conducting materials, radio-frequency (rf) skin depth is a problem. For materials with large dimensions compared to skin-depth, the rf field does not penetrate completely and hence the signal will not be observed or attenuated strongly. The skin-depth issue is addressed in more detail in Chapter 6.
1.6 Goal and Scope of this thesis
This thesis describes the NMR investigation of siting and mobility of hydrogen in Mg-based hydrogen storage materials. Most studies on Mg-Mg-based hydrides utilize diffraction-Mg-based techniques and/or Sievert methods to obtain structural, kinetic and thermodynamic parameters. NMR has the advantage that the chemical location of hydrogen as well its mobility can be monitored directly. This will give an atomistic view of the energy carriers, i.e.the hydrogen atoms and also their mobility within in the energy storage material. In our MAS NMR study of Mg-TM-hydrides, the primary goal is to distinguish between hydrogen atoms with different metal coordination, Qn = H-MgnTM4-n, where n is the number of Mg atoms in the first
coordination sphere. Do hydrogen atoms from one metal-coordination environment move to all other coordination environments? How are the different metal coordinations arranged within the lattice? What is a typical length scale separation between different metal coordinations?
Chapter 1 gives a general introduction to the thesis. The importance of the hydrogen economy is discussed. The criteria for a material to be considered for hydrogen storage are elucidated. General overviews of different types of materials that are currently studied for hydrogen storage are mentioned. We narrow down to MgH2 and other Mg-based hydrogen
storage materials, which this thesis about. A short overview is given about NMR methodology and its advantages to study metal hydrides. The aim of the thesis and important results are described in each chapter.
Chapter 2 gives an overview of the theory behind the characterization techniques. NMR is used extensively in this thesis and therefore more emphasis is given for various pulse sequences that are used to study the Mg-TM based hydrogen storage materials.
.
Chapter 3 describes the nano-structure of Mg0.65Ti0.35D0.65 synthesized by ball-milling Mg
and Ti and subsequent gas phase deuterium loading at high pressure and temperature. MAS NMR spectroscopy, powder X-ray diffraction (XRD) and neutron diffraction (ND) were used to study the complex structure of the material. Deuterium dynamics were studied by use of 2D Exchange NMR spectroscopy (2D Exsy) at room temperature and by use of 1D Exchange NMR spectroscopy (1D Exsy) at various temperatures. XRD and ND indicate macro-phase separation of MgTi alloy into MgD2 and TiD2 after deuteration. However, NMR shows the presence of an
additional TiDy nanophase which is XRD-invisible. Further, 2D Exsy NMR reveals deuterium
exchange between this XRD invisible TiDy nanophase and MgD2. Comparing the intrinsic cell
parameters of rutile MgH2 and fluorite TiH2, we propose that stabilization of the mixed
nanocomposite may arise from a coherent coupling between the crystal structures of the rutile MgD2 nanodomains and the thin layers of fcc TiDy.
Chapter 4 deals with Mg0.65Ti0.35D1.1, where first the MgTi alloy is synthesized as a film
by magnetron sputtering and, subsequently, deuterated by means of gas absorption at room temperature. For this co-sputtered material NMR shows no macro-phase separation into MgD2
and TiD2, unlike for the ball-milled Mg0.65Ti0.35D0.65 material discussed in the preceding chapter.
It does reveal the presence of deuterium in Mg-rich and Ti-rich coordination states. Interestingly, 2D Exsy shows that most of the of deuterium atoms within Ti-rich clusters and Mg-rich clusters exchange with one another. A minor fraction, which does not participate in the exchange process appears stably bound to titanium. The observed deuterium exchange and the reduced Knight shift compared to bulk TiD2 are explained using a model with TiD2 nanoslabs.
In general NMR is considered to be quantitative. This is very much important in the present study as we quantify the energy carriers.
Chapter 5 deals with the NMR visibility of deuterium atoms in the melt-cast Mg0.65Sc0.35D2.2. By use of 2H-{45Sc} Double Quantum spectroscopy, we were able to distinguish
between deuterium in Mg-rich and Sc-rich sites. 1D Exsy reveals that total peak area is not conserved during deuterium exchange between the Mg-rich and Sc-rich sites which indicates the presence of an NMR-invisible or “dark” deuterium fraction. As possible origins of the reduced deuterium visibility deuterium mobility at the sample-rotation timescale and second-order quadrupolar line broadening are discussed.
Chapter 6 discusses hydrogen storage in Mg melt-infiltrated nanoporous carbon at
different MgH2: carbon ratio. The complex nanostructure of the material is characterized by a
combined NMR, XRD and Temperature Programmed Desorption(TPD) approach. The conductive nature of nano-porous carbon reduces the NMR visibility of the hydrogen atoms. The variation of NMR signal intensity with the packing density of the material is probed to estimate the NMR visibility in these materials. MgH2 within the pores of carbon is not detected by XRD.
nanophase inside the pores of the carbon support and bulk MgH2 outside the carbon on the basis
of spin-lattice relaxation, motional lineshape narrowing in static NMR and chemical-shift differences in MAS NMR.
Chapter 7 gives the summary of the present research work.
1.6 References
1J. M. M. Amouroux, IEPE, 2003.2
L. Schlapbach. and A. Züttel, Nature, 414, 353, 2001.
3
R. Coontz and Brooks Hanson, Science, 305, 957, 2004.
4
R. F. Service, Science, 305, 957, 2004.
5
L. Barreto, A. Makihira and K. Riahi, Int J Hydrogen Energy , 28, 267, 2003.
6
DOE Hydorgen, Fuelcells and infrastructure Technoligies Program, Multi-Year Research, Development and Demonstration Plan: Planned Program Activities for 2005-2015, http://www1.eere.energy.gov/hydrogenandfuelcells/mypp/index.html, 2009.
7
A. C. Dillon, K. M. Jones, A. Bekkeddahl, C. H. Kiang, D. S. Bethune, M. J. Heben, Nature,
386, 377, 1997.
8
M. Hirscher, M. Becher, M. Haluska, U. Dettlaff-Weglikowska, A. Quintel, G. S. Duesberg, Y. M. Choi, P. Downes, M. Hulman, S. Roth, I. Stepanek and P. Bernier, Appl. Phys. A, 72, 129, 2001.
9
H. Kajiaura, S. Tsusui, K. Kadono, M. Kakuta, M. Ata and Y. Murakami, Appl. Phys. Lett., 82, 1105, 2003.
10
A. R. Harutyunyan, B .K Pradhan, T. Tokune, Y. Fujiwara and P. C. Eklund, in Proceedings of the 25th International Conference on Carbon, CARBON ‘01, American Carbon Society, July 14–19 Lexington, KY, 2001.
11
N. L. Rosi, J. Eckert, M. Eddaoudi, D. T. Vodak, M. O’Keefe and O. M. Yaghi, Science, 300, 1127, 2003.
12
G. F´erey, M. Latroche, C. Serre, F. Millange, T. Loiseau, A. Percheron- Gu´egan, Chem.
Comm., 24, 2976, 2003.
13
L. Pan, M. B. Sandler, X. Huang, J. Li, M. Smith, E. Bitther, B. Bockrath and J. K. Johnson,
14
J. H. N. van Vught, F. A. Kuijpers and H. C. A. M. Bruning, 25, 133, 1970.
15
J. J. Reiley and R. H. Wiswall, InorgChem, 13, 218, 1974.
16
P. H. L. Notten and M. Latroche, Encyclopedia of Electrochemical Power Sources, 4, 502–521, 2009.
17
B. Bogdanovic and M. Schwicardi, J. Alloys Comp., 253, 1, 1997.
18
B.C. Hauback, Z. Kristallogr., 223, 636, 2008.
19
S. Orimo, Y. Nakamori, J. R. Elisco, A. Zuttel and C. M. Jensen, Chem. Rev., 107, 4111, 2007.
20
P. Chen, Z. T. Xiong, J. Z. Luo, J. Y. Lin and K. L. Tan, nature, 420, 303, 2002.
21
A. Züttel, Materialstoday, 18, 2003.
22
W. H. Zachariasen, C. E. Holley, Jr., and J. F. Stamper, Jr., Acta Crystallogr.
16, 352, 1963.
23
P. H. L. Notten, M. Ouwerkerk, H. van Hal, B. Beelen, W. Keur, J. Zhou and H. Feil, J. Power
Sources, 129, 45, 2004.
24
R. A. H. Niessen and P. H. L. Notten, J. Alloys Compd., 404-406, 457, 2005.
25
W. P. Kalisvaart, R. A. H. Niessen and P. H. L. Notten, J. Alloys Compd., 417, 280, 2004.
26
M. Latroche, W. P. Kalisvaart and P. H. L. Notten, J. Solid State Chem., 179, 3024, 2006.
27
M. S. Conradi, M. P. Mendenhall, T. M. Ivancic, A. E. Carl, C. D. Browning, P. H.L. Notten, W. P. Kalisvaart, P. C. M. M. Magusin, R. C. Bowman Jr., S. J. Hwang and N. L. Adolphi, J.
Alloys Compd., 446–447, 499, 2007.
28
L. Vegard, Z. Physik, 5, 17, 1921.
29D. Kyoi, T. Sato, E. Ronnebro, N. Kitamura, A. Ueda, M. Ita, S. Katsuyama, S. hara, D. Noreus and T. Sakai, J. Alloys Compd., 372, 213, 2004.
30
D. Kyoi, T. Sato, E. Ronnebro, N. Kitamura, A. Ueda, M. Ita, S. Katsuyama, S. hara, D. Noreus and T. Sakai, J. Alloys Compd., 375, 253, 2004
31
D. Mosier, DF. J. Bull, T. Sato, D. Noreus, D. Kyoi, T. Sakai, N. Kitamura, H. Yusa, T. Taniguchi, W. P. Kalisvaart and P. HJ. L. Notten, J. Mater. Chem., 19, 8150, 2009.
32
T. Sato, D. Kyoi, E. Ronnebro, N. Kitamura, T. Sakai and D. Noreus, J. Alloys Compd.,417, 230, 2006.
33
V. O. Hellenis, H. Kandler, E. Leicht, W. Quiring and E. Wolfel, Z. Anorg. Alleg. Chem., 320, 86, 1963.
34
J. Huot, G. Liang, S. Boilya, A. Van Nesteb and R. Schulz¸ J. Alloys Compd., 293, 495, 1999.
35
A. Zaluska, L. Zaluski, and J.O. Strom–Olsen, J. Alloys Compd, 288, 217, 1999.
36
H. G. Schimmel, M. R. Johnson, G. J. Kearley, A. J. Ramirez-Cuesta, J. Huot and F. M. Mulder, Mater. Sci. Eng. B, 108, 38, 2004.
37
H. G. Schimmel, M. R. Johnson, G. J. Kearley, A. J. Ramirez-Cuesta, J. Huot, and F. M. Mulder, J. Alloys Compd., accepted for publication.
38
G. Lianga , J. Huot, S. Boily, A. Van Neste and R. Schulz, J. Alloys Compd., 292, 247, 1999.
39
G. Barkhordarian, T. Klassen and R. Bormann, J. Alloys Compd. 364, 242, 2004.
40
G. Barkhordarian, T. Klassen and R. Bormann, Scripta Mater. 49, 213, 2003.
41
R. W. P. Wagemans, J. H. van Lenthe, P. E. de Jongh, A. J. van Dillen and K. P. de Jong, J.Am.
Chem.Soc., 127, 16675, 2005.
42
P.E. de Jongh, R.W.P. Wagemans, T.M. Eggenhuisen, B.S. Dauvillier, P.B. Radstake, J.D. Meeldijk, J.W. Geus and K.P. de Jong, Chem. Mater. 19, 6052, (2007).
43
S. Singh, PhD Thesis, Delft University of Technology.
44
N. Iqbal, N. H. van Dijk, S. E. Offerman, M. P. Moret, L. Katgerman and G. J. Kearley, Acta
Mater., 53, 2875, 2005.
45
T. J. Richardson, R. D. Armitage, J. L. Slack and M. D. Rubin, poster presentation at the Fourth International Meeting on Electrochromism (IME-4), August 21-23, Uppsala, Sweden.
46
T. J. Richardson, R. D. Armitage, J. L. Slack, R. Kostecki, B. Farangis and M. D. Rubin, Appl.
Phys. Lett., 78, 3047, 2001.
47
M. Slaman, B. Dam, M. Pasturel, D. M. Borsa, H. Schreuders, J. H. Rector and R. Griessen,
Sens. Actuators B, 123, 538, 2006
48
D. M. Borsa, A. Baldi, M. Pasturel, H. Schreuders, B. Dam, R. Griessen, P. Vermeulen and P. H. L. Notten, Appl. Phys. Lett., 88, 241910, 2006.
49
S. Bao, K. Tajima, Y. Yamada, M. Okada and K. Yoshimura, Appl. Phys. A, 87, 621, 2007.
50
A. Baldi, D. M. Borsa, H. Schreuders, J. H. Rector, T. Atmakidis, M. Bakker, H. A. Zondag, W. G. J. van Helden, B. Dam and R. Griessen, Int. J. Hydrogen Energy, 33, 3188, 2008.
51
E. G. Maksimov and O. A. Pankratov, Usp. Fiz. Nauk, 116, 385, 1975.
52
K. Morimoto, M. Saga, H. Fujii, T. Okamoto and T. Hihara, J. Phys. Soc. Jpn., 57, 647, 1988.
53
54
U. Eberle, G. Majer, A. V. Skripov and V N Kozhanov, J. Phys.: Condens. Matter, 14 153– 164, 2002.
55
J. Gottwalda, G. Majer, D.T. Petersonc and R.G. Barnesc, J. Alloys Comp., 356–357, 274, 2003.
Chapter 2
Characterization Techniques
2.1 Introduction
A single characterization technique alone will not give a complete picture of heterogeneous materials. Every technique has its own advantages and disadvantages. Diffraction-based techniques such as powder X-ray diffraction, neutron diffraction are commonly employed to study the crystalline phase composition of metal hydrides. However, these methods require long-range ordering of the atoms in materials. The long-range ordering is not necessary for Nuclear Magnetic Resonance (NMR) spectroscopy. The disadvantage is that unhydrogenated portions of a metal hydride cannot be detected with 1H NMR. Therefore, a combination of characterization techniques to exploit the advantages of every technique is a must to get a full picture of materials. This chapter describes the theory behind the techniques that are used to characterize the metal hydrides. A major part of the work described in this thesis involves solid state Nuclear Magnetic Resonance (NMR). This will be dealt with in a more detailed way. In the last section, basic aspects of powder X-ray diffraction are given.
2.2 Nuclear Magnetic Resonance (NMR)
NMR is a tool, which is widely used in many disciplines of fundamental and applied sciences. Most often NMR is used as a structural characterization tool, mobility can also be probed. NMR involves resonance absorption of electro-magnetic energy by nuclear moments. This phenomenon is observed for atomic nuclei which have non-zero spin I. These nuclei have magnetic dipole moments given by
I
µ=γ h (2.2.1) where γ is the gyro-magnetic ratio of the nucleus, and I = (Ix, Iy Iz) the spin-operator vector
composed of the three spin operators Ix, Iy and Iz. As a result if their magnetic momnent atomic
nuclei interact with a magnetic field B0. The corresponding Zeeman interaction is given by the
Hamiltonian: 0 0 . B I B µ • =− • − =
γ
h z H (2.2.2)The eigen values of this Hamiltonian represent the so-called Zeeman energy levels. A nucleus with spin I has |2I + 1| energy levels associated with equally many spin angular momentum projections on the magnetic-field axis. A spin ½ nucleus has two energy levels, separated by ∆E = γB0 = hγω0, ω0 is the larmor frequency. This directly indicates that sensitivity of NMR signal is
enhanced with increasing magnetic field.
The basic nuclear spin interactions in solid-state NMR,1,2 are: (i) Chemical shift interaction
(ii) Dipole-dipole interaction (iii) J- coupling interaction (iv) Quadrupole interaction
(i) Chemical shift interaction. The magnetic field experienced by the nucleus is shielded by the surrounding bonding electrons. The chemical shielding Hamiltonian acting on a spin I is
0 B σ I
γ
h − = z H (2.2.3)where σ is the second rank tensor describing the chemical shielding. The chemical shieldingcan be separated into isotropic and anisotropic components. The isotropic component is
z iso cs B I H ( ) 3 1 33 22 11 0 σ σ σ γ + + − = h (2.2.4)
and the anisotropic component is
z z aniso cs I I H [( )sin cos2 ] 2 1 ) 1 cos 3 )]( ( 2 1 [ 3 1 2 22 11 0 2 22 11 33 0 σ σ σ β ϖ σ σ β γ ϖ − + − − − − = (2.2.5)
where σ11, σ22 and σ33 are the principal axis values of the chemical shielding tensor, and the
angles β and γ relate the Principal Axes System (PAS) of the chemical shielding tensor to the lab frame in which the magnetic field is oriented along the z axis.
If the chemical shift anisotropy, ∆σ, is defined as σ33 – σiso and the chemical shift asymmetry, η as
(σ22 – σ11)/ ∆σ, the chemical shielding Hamiltonian can be rewritten as: z z
iso
cs I I
H [(3cos 1) sin cos2 ]
2 1 2 2 0 0σ ω σ β η β α ω − ∆ − + − = h h (2.2.6)
(ii) Dipolar interaction. In the strong field approximation, the truncated form of homonuclear dipolar Hamiltonian representing the magnetic dipole coupling between two spins is given by
) . 3 )( (cos (1) (2) (1) (2) 2 −I I − = D z z d P I I H ω θ (2.2.7)
where ωD is the dipolar coupling constant given by
= 3 0 4 r S I D h γ γ π µ ω (2.2.8)
with distance r. P2(cosθ) is the second order Legendre polynomial given by
) 1 cos 3 ( 2 1 ) (cos 2 2 θ = θ − P (2.2.9)
with θ the angle between the internuclear vector connecting the two spins and the external magnetic field B0.
(iii) J-coupling interaction. Indirect spin-spin interaction which is also called J-coupling interaction is the interaction between nuclei mediated through the bonding electrons in the molecule. The Hamiltonian is given by
) ( (1) I J I . . 2 = J H (2.2.10)
where J is a second-rank tensor.
(iv) Quadrupolar interaction: Nuclei with spin I > ½ have electric quadrupole moment which interacts with the electric filed gradients from surrounding charges. In a strong magnetic field the quadrupolar interaction can be treated as a perturbation to the Zeeman interaction and can be expressed as a sum of first- and second-order terms.
HQ = HQ1 + HQ2 (2.2.11)
The first-order quadrupolar Hamiltonian is given by
[
(3 ( 1) ( )]
) , ( 6 1 2 2 2 1 q z x y Q I I I I I H = hω β γ − + +η − (2.2.12)where ωq (β,γ) is the orientation dependent quadrupolar frequency given by
) 2 cos sin 1 cos 3 ( 2 ) , (β γ ω 2β η 2 β γ ωq = Q − + (2.2.13)
where β and γ denote the polar angles defining the orientation of the quadrupolar tensor V in the laboratory frame. The qaudrupolar constant ωQ and asymmetry parameters η are given by
zz Q V I I qQ e h ) 1 2 ( 2 3 2 − = ω (2.2.14) and zz yy xx V V V − = η (2.2.15)
whereby Vzz, Vyy and Vxx are the principal axis values of the quadrupolar tensor V.
The first-order quadrupola shift of the unpertubed Zeeman levels is proportional to (3m2 –1), where m is the magnetic quantum number associated with a particular Zeeman level. Therefore, the frequency of the central transition - ½ ↔ ½ of half integer quadrupolar nuclei, like 27Al or
45
Sc, or, in general, “symmetric” transition –m ↔ m is not affected by the first-order interaction. For large quadrupolar coupling interactions, the second-order energy correction has to be considered. The second-order quadrupolar Hamiltonian is of the form:3
[
3]
0 2 ) 2 ( ) , ( ) , ( z z Q Q A I B I Hβ
γ
β
γ
ω
ω
+ − = (2.2.15)where A(β,γ) and B(β,γ) are orientation dependent geometrical factors.The second-order term is inversely proportional to the Larmor frequency. Thus the significance of this term becomes less at increasing magnetic field.
2.2.1 Spectral editing
A challenge for most of solid-state NMR spectroscopist is to obtain a liquid like spectrum for solid samples. In solids, broad NMR lines are caused by Chemical Shift Anisotropy(CSA), dipole or quadrupolar interactions. The nuclear spin Hamiltonian consists of a spatial and spin part. The spatial part can be averaged by the mechanical rotation of the samples. This gives a narrow signal with spinning sidebands. The spin part can be dealt by applying radio-frequency (RF) pulses sequences, which allow us to obtain specific information, for example, forbidden transitions corresponding to multiple quantum transition can be observed in an indirect two-dimensional method. For spin 5/2 nuclei, triple quantum spectrum gives narrower linewidth than the single quantum spectrum. In this section, various pulse sequences which are mentioned in this thesis to study the siting and dynamics of hydrogen atoms are described.
2. 2.1.1 Magic Angle Spinning
Magic Angle Spining (MAS)4,5 is often employed to get higher chemical resolution. This involves rotating the sample at a angular velocity ωr around an axis, inclined at a particular angle
β with respect to the magnetic field. Fig. 2.1 shows the description of the same. rij is the
inter-nuclear vector connecting the two spins i and j. θij(t) is the angle that rij makes with the applied
magnetic field, H0. The angle θij representing the angle between rij and magnetic field, becomes
time dependent given by
cosθij(t) = cosβcosβ`ij + sinβsinβ`ijcos(ϕij + ωrt) (2.2.1.1.1)
The β`ij is the angle the rotor makes with the inter nuclear vector rij. After substituting Eq 2.2.17,
the cos2θ term in Eq. 2.2.9, we get,
½(3cos2β -1 )(3cos2β`ij – 1) + 3/2sin2βsin2β`ij(ϕjk + ωrt) + 3/2sin2βsin2β`ijcos2(ϕjk + ωrt)
(2.2.1.1.2)
Figure 2.1 Diagram showing the rotation of internuclear vector rij when the sample is rotated at an angle β
with respect to the direction of applied magnetic field.
The first term vanishes for β = acos( 1/3) = 54.74° (magic angle) and hence gives a narrow resonance. The second and third term is periodic with ωr and gives rotation sidebands at
2.2.1.2 Deuteration Effect
Most of the experiments described in this thesis are performed on deuterated compounds. The dipolar coupling constant for protons separated by a distance 1
Å
can be calculated with Eq 2.2.8. This corresponds to ~120 kHz. As mentioned before, MAS is often employed to get narrower lines. However, even the state-of-art ultrafast MAS upto 70 kHz will not be able to average dipolar coupling completely. Another way to get narrow lines is with deuterium NMR. The gyro-magnetic ratio of deuterium (2H) is ~ 1/6.5 less than that of hydrogen (1H) . Hence, the dipolar coupling constant for two deuterium atoms separated by 1Å
, is ~ 2.8 kHz. This can be easily averaged with relatively moderate spinning rates.2.2.1.3 TRAPDOR (TRAnsfer of Population in DOuble Resonance)10
The heteronuclear dipolar coupling between two spins, I and S, is probed by use of this pulse sequence. In the strong-field approximation the heteronuclear dipolar Hamiltonian is given by ) )( (cos 2 z z D d P I S H =−ω θ (2.2.1.3.1)
Fig. 2.2 shows the TRAPDOR pulse sequence. The nucleus I is observed with and without irradiation at the resonance frequency of S spins as a function of irradiation time.
Figure 2.2 (a) Hahn echo pulse sequence (b) TRAPDOR with irradiation on S spins. The time interval, τ,
takes a value nτMAS, where n is an integer and τMAS is the time taken by the rotor for one complete rotation.
During MAS, the orientation of the dipole and quadrupolar tensors becomes time dependent. The combined action of MAS as well as rf irradiation on S spins results in exchange of population of the Zeeman states of the S spins. This has a clear effect on dipolar coupling
I I
S
τ τ τ τ
between the spins I and S. The Hahn-echo signal with S spin recoupling decays faster as a function of the echo time 2τ than the signal without irradiation on the S spins.
2.2.1.4 Deuterium Double Quantum (DQ) NMR
Deuterium nuclei have spin 1 and therefore there are three energy levels in a strong magnetic field. The double-quantum (DQ) transition –1 ↔ 1 is independent of quadrupolar coupling to first order. This is illustrated in Fig. 2.3 Therefore, narrow lines are expected in a double quantum spectrum.1112 As a novel approach, we have irradiated on 45Sc nuclei during the DQ evolution time t1. Fig. 2.2.1.4.2 shows the pulse sequence and the coherence pathway
diagram13 used to record the two-dimensional DQ spectrum.
Figure 2.3 The energy levels of spin 1 nucleus. The observable single-quantum transitions are affected by
the first-order quadrupolar interaction, whereas, the double-quantum coherences are not affected.
Figure 2.4. (a) Pulse sequence used to record Two-Dimensional (2D) Double Quantum MAS NMR spectra.
The pulse length of all pulses are π/2. 45Sc is irradiated during the evolution period.
+2 +1 p = 0 -1 -2 t1 = nτMAS τex τex τz t2 υ0 υ0 υ0 + 1/2 ∆υq υ0 - 1/2 ∆υq I = 1 1 0 -1 Zeeman interaction Effect of first-order quadrupolar interaction 2υ0 2 H 45 Sc
The DQ coherences are excited by first pair of pulses and reconverted to longitudinal magnetization by the second pair of pulses. The time interval, τZ, has negligible duration and it is
present to allow a clean phase shift of the pulses. In our DQ MAS NMR experiments we have used rotor synchronization in the indirect dimension by increasing the evolution time t1 in steps
of the sample rotation time T1. The States-Haberkorn-Ruben method is used for sign
discrimination in F1 dimension and pure-absorption 2D spectrum is obtained by hypercomplex Fourier transform.
2.2.2 Mobility from NMR/ Dynamics
NMR is a powerful technique to probe the motions in solids. The motion of hydrogen atoms in metal hydrides can be monitored. By studying the rate of the motions as a function of temperature, we can experimentally probe the activation barriers. This is crucial if we are to fundamentally understand the structure and properties of hydrogen-storage materials. This gives scope for further improvement of materials that are used in the battery electrode.
Different NMR experiments are employed to probe the dynamics at various time scales. Motions with correlation time (τc) in the order of 10-2 s or longer are best studied with
Two-Dimensional exchange spectroscopy (2D Exsy).14,15 Motions with rates τc-1 of the order of the
nuclear spin interaction anisotropy can be accessed via lineshape analysis.16,17 Motions in the order of 10-6 to 10-9 s are out of the dynamic range of lineshape analysis. Spin-lattice relaxation is investigated in such circumstances. Thus, in cases where relaxation is dominated by one particular nuclear spin interaction, the spin-lattice relaxation times can be calculated for different motions and compared with experimental values to reveal motional details.18,19 Molecular mobility regimes and the corresponding NMR experiments to observe the same are summarized in Table 2.2.2.1.
Table 2.2.2.1. Motional regimes which can be probed by various NMR techniques are given below.
NMR technique Correlation time (s) 2D Exsy/1D Exsy 10-2- 101 Lineshape analysis/T2 10-2 – 10-4
T1ρ 10-4 – 10-5
T1 10-6-10-9
2.2.2.1 Two- Dimensional Exchange NMR spectroscopy (2D Exsy)
The hydrogen motion between NMR distinct sites is probed by this experiment. Hydrogen or deuterium exchange processes between two NMR distinct states a and b can be classified into three categories: slow, intermediate and fast exchange depending on the exchange rate k and the frequency separation |νa – νb| between NMR frequencies νa and νb of sites a and b.
In the slow-exchange regime, k << |νa – νb|, the NMR lineshapes are not broadened significantly.
For fast exchange, k >> |νa – νb|, a single signal is observed at a weighed-average frequency, ν =
Pa νa + Pb νb where Pa and Pb are the population of each site with Pa + Pb = 1. In the intermediate
exchange regime, k ~ |νa – νb|and the two NMR lines coalesce into a broad line. 2D Exsy is a
powerful technique to observe slow exchange processes.
In general, all 2D experiments consist of a preparation, evolution, mixing and detection period. Fig. 2.5 shows the typical pulse sequence of a 2D Exsy experiment. In the preparation period, the first pulse creates the transverse magnetization. During the evolution time, t1, the
magnetization vectors of the individual resonances precess at their characteristic resonant frequencies. Hence, initial resonances are frequency-labeled during t1. At the end of t1, the
second pulse flips the x or y magnetization component onto the longitudinal axis and the mixing period starts. It is in this period that spins have an opportunity to move to a chemical distinct site. The final pulse generates the observable signal during the detection period, t2. The 2D spectrum
is obtained by Fourier transform with respect to t1 and t2. The exchange between spins from
that do not exchange. Complete exchange of spins from chemical distinct sites is indicated by the similarity of the horizontal or vertical trace and the horizontal or vertical projection onto the direct or indirect frequency axis, respectively. The mixing period, tm, plays a crucial part in a 2D
experiment as this is the time during which the exchanges are monitored. Very low values of tm
compared to exchange time will result in diagonal peaks only and very high values compared to spin-lattice relaxation time (T1) will result in a weak signal.
Figure 2.5: Pulse sequence used in a typical 2D Exsy experiment. The pulse widths of all pulses are π/2.
The 2D spectrum can be amplitude modulated or phase modulated. The amplitude-modulated spectrum gives pure absorption lineshapes where as modulated spectrum results in phase-twisted lineshapes. A typical 2D Exsy spectrum is illustrated in Fig. 2.6. It concerns a 2H MAS NMR spectrum recorded for ball-milled Mg0.65Ti0.35D0.65 which is the topic of Chapter 3. Three
signals A,B and C can be recognized in the 1D spectrum and projection above the 2D spectrum, a well as along the spectral diagonal. The cross peaks between A and B denote deuterium exchange of spins between the corresponding environments. The interpretation of the resulting 2D spectra is essentially model free. However, it tends to be time consuming. E.g., for the materials investigated in our study, it takes approximately 18 hrs for a single experiment. The exchange rate can also be estimated in a faster way from one-dimensional exchange spectroscopy (1D Exsy), which is explained in the next section.
Figure 2.6 (a) 1D 2H NMR spectrum of ball-milled Mg0.65Ti0.35D0.7. The three signals labeled A, B and C are
assigned to deuterium atoms at three inequivalent sites. (b) 2D Exsy spectrum: The crosspeaks between A and B indicate that there is exchange of deuterium atoms between A and B sites.
2.2.2.2 One Dimensional Exchange Spectroscopy (1D Exsy)
As mentioned in sec 2.2.2.1, 1D Exsy is used for obtaining the exchange rate constant. In this technique, the polarization of hydrogen atoms from one of the sites is selectively perturbed followed by a non-selective pulse. As a result of hydrogen exchange during the following mixing time, hydrogen atoms with perturbed polarization will replace hydrogen atoms at the “unperturbed” sites and vice versa. As a consequence, the signal intensity of the originally non-perturbed sites will decrease, and that of the initially non-perturbed sites will increase as function of the mixing time. The longest exchange timescales that can be probed by this method is determined by spin-lattice relaxation. Even without deuterium exchange, any perturbed spin polarization of the deuterium nuclei will relax to the thermal equilibrium value.
The polarization of deuterium atoms from site A (Fig. 2.6 a) is selectively perturbed and monitored as a function of time. Fig. 2.7 shows the polarization of deuterium atoms from A, B and C sites. The polarization of the B site decreases initially and then
B C 2H NMR shift (ppm) (a) (b) -300 -100 500 300 100 0 A
0 10 20 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 A B C M ag n e ti za ti o n ( a .u ) Time (s)
Figure 2.7. Polarization of deuterium atoms as a function of time from distinct sites A, B, C after
perturbation of only the A site.
spin-lattice relaxation dominates. Magnetization recovery of site A shows bi-component behavior in time. The fast component is related to the exchange and the slow component to spin-lattice relaxation time (T1). The polarization of deuterium atoms from the site C remains
constant, which indicates that deuterium atoms from this site do not take part in the exchange with A or B site. This is also seen in the 2D Exsy spectrum in Fig.2.6b, where there are no crosspeaks between site C and other sites.
The 1D Exsy is performed at different temperatures and the exchange rate is monitored as a function of the same. This gives the activation barrier for the deuterium atom exchange between A and B sites.
There are different ways to achieve selective perturbation of polarization of the nucleus. One way is to apply a long soft pulse at a desired resonance frequency. In another approach, shaped pulses are applied. A Gaussian shaped pulse is considered as an example. This has a disadvantage that amplitude is not constant throughout the width of the resonance. This will lead to phase distortion of the signal.
In another method to selectively perturb a resonance line, DANTE20 pulse sequence is employed. DANTE is an abbreviation for Delays Alternating with Nutations for Tailored
Excitation. This consists of series of mRF pulses separated by time, τ. The overall duration of the pulse sequence is ∆t. Upon Fourier transformation, the excitation pulse in the frequency domain consists of series of pulses of width 1/∆t Hz, separated by 1/ τ Hz. The RF carrier should be positioned at the desired frequency of perturbation. The number of pulses (m) and power of the pulse can be varied to achieve a certain flip angle for a particular resonating signal. For MAS experiments, the value of τ = τMAS is chosen in such way so as to perturb the centerband and the
sidebands (if present) of the signal in a similar way. The time as well as frequency domain spectrum of excitation pulses are shown in the Fig. 2.8.
Figure 2.8. Selective Excitation using DANTE pulse sequence. (a) Time domain pulse sequence. (b) Fourier
transformation of the excitation pulse.
Both 2D and 1D Exsy techniques are powerful tools to monitor the motion in millisecond to a second timescale. However, each technique has its advantages and limitations..
2.2.2.3 Relaxometry
2.2.2.3.1 Spin-lattice relaxation
This process probes the mobility in nanosecond time scale. There are different techniques to extract the spin-lattice relaxation time (T1) namely inversion recovery, saturation recovery,
progressive saturation. The saturation recovery technique (SR) will be discussed since this τ
∆t
(a)
technique has been used it in this work. The pulse sequence for SR is shown in the Fig. 2.9 (a) The intensity of the signal is given by
S(τ) = M0(1 - exp(-τ/T1)) (2.2.2.3.1.1)
Fig 2.9 (b) shows the signal intensity as a function of time.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 2 4 6 8 Time (s) M a n e ti z a ti o n ( a .u )
Figure 2.9 (a) Pulse sequence for saturation recovery. (b) Plot of signal intensity as a function of time using
the SR technique.
2.2.2.3.2 Spin-lattice relaxation in rotating frame
The atomic motions in the order of kHz are monitored via this technique. Fig. 2.10 shows the pulse sequence to measure the spin-lattice relaxation in rotating frame (T1ρ). The first pulse rotates the magnetization from equilibrium to x-y plane. The spinlock pulse is applied along x or y direction. A typical strength of spinlock field (ω1) is ~ 50- 100 kHz. The intensity of the signal
as a function of spinlock time is given
S(τ) = M0 exp(-τ/T1ρ) (2.2.2.3.2.1)
Typically, T1 and T1ρ are measured as a function of temperature and their values goes through a
minimum. The minimum happens when the product of the Larmor frequency (ω0) or the spinlock
field (ω1) and the motion correlation time (τc) is roughly equal to 1.
Figure 2.10 Pulse sequence for measuring spin lattice relaxation time in rotating frame. The signal intensity
is monitored as a function of spinlock time τ.
2.3 X-ray diffraction (XRD)
XRD21 technique is commonly employed to determine the crystallographic phases present in a material. A lattice consists of periodic arrangement of atoms with lattice parameters a, b, and c representing the lengths in three different orthogonal axis. In three dimensions, the lattice points form into set of planes separated by a distance called as interplanar distance, d. There can be different sets of planes with distinct d values. The rays are scattered by the electrons from the atoms. The scattered rays from different planes undergo constructive interference to give a peak. The condition for interference is given by Bragg equation,
nλ = 2 d sinθ (2.3.1)
where n is an integer known as order of diffraction, λ is the wavelength of x-rays and θ is the angle of incidence.
θ
θ
θ
d
Fig 2.11. Illustration of X-ray diffraction. θ is angle of incidence of the X-rays. The distance between atomic planes is given by d.
Miller indices are used to describe the planes in the crystal. They are represented by (hkl). The distance between two planes (dhkl) and the miller indices is related to the crystal
dimension. For simple cubic lattice, τ
) (h2 k2 l2 a dhkl + + = (2.3.2)
The width of the diffraction peak gives an estimate about the crystallite size. The crystallite size is given by
θ β λ cos K D= (2.3.3)
Where D is the diameter of the crystal, K is the (typical value is 0.9) shape factor, λ is the wavelength of the incident X-ray beam, β is the full-width at half-maximum (FWHM) measured in radians and θ is the Bragg angle.
2.4 References
1
Abragam A, Principles of Nuclear Magnetism (Oxford: Oxford University Press). 1983
2
Slichter C. P., Principles of Magnetic Resonance, (Berlin: Springer), 1990.
3
Iuga, D., Nuclear magnetic resonance studies of half-integer quadrupolaar nuclei, Thesis (Radboud University 2003)
4
Andrew E. R, Bradbury A, Eades R. G., Nature, 183,1802, (1959).
5
Lowe I. J., Phys. Rev. Lett., 2,285, (1959).
10
E.R.H. Van Eck, R. Janssen, W.E.J.R. Maas, W.S. Veeman, Chem. Phys. Lett. 174 428, (1990).
11
S. Vega and A. Pines, J. Chem. Phys., 66, 5624, (1977)
12M. Cutajar, S.E. Ashbrook, S. Wimperis, Chem. Phys. Lett. 423, 276, (2006)
13
G. Bodenhausen, H. Kogler, R.R. Ernst, J. Magn. Reson. 58, 370, (1984).
14
C.L. Perrin, J. Magn. Reson. 82, 619, (1989).
15A.D. Bain, J.S. Martin, J. Magn. Reson. 29 (1978) 125
16
T. Fauconnier, C.J.L. Lock, R.A. Bell, J.F. Britten, K. Rainsford, Can. J. Chem. 72 382, (1994)
17
C. Deverell, R.E. Morgan, J.H. Strange, Mol. Phys. 18, 553, (1970)
18
R. C. Bowman. Jr., A. J. Maeland , W. K. Rhim, Phys. Rev. B., 26, 6362, (1982)
19U. Eberle, G. Majer, A. V. Skripov, V. N. Kozhanov, J. Phys.: Condens. Matter 14 153–164, (2002)
20
Morris G.A. and Freeman R., J.Mag. Reson.,29, 423, (1988).
21
Chapter 3
Nanostructures of Mg
0.65Ti
0.35D
xstudied with X-ray
diffraction, neutron diffraction, and magic-angle-spinning
2
H NMR spectroscopy
*Abstract
Magnesium transition-metal alloys have a high hydrogen-storage capacity and show improved hydrogen uptake and -release kinetics compared to pure magnesium. In the present study we have investigated the structure of bulk magnesium-titanium deuteride Mg0.65Ti0.35Dx prepared via
mechanical alloying and gas-phase deuterium absorption by combined use of x-ray diffraction (XRD), neutron diffraction, and magic-angle spinning 2H nuclear magnetic resonance (NMR) The initial ball-milled alloy has two XRD-distinct Mg and Ti fcc phases. Even after prolonged exposure to deuterium gas at 75 bar and 175 °C the materials with and without palladium catalyst are only partly deuterated. Deuterium loading causes the formation of, on the one hand, bct (rutile) MgD2 nanodomains with interdispersed TiDylayers and, on the other hand, a separate
fcc (fluorite) TiDzphase. The TiDy phase is XRD invisible, but shows clearly up at a 2H NMR
shift of −43 ppm between the shift of MgD2 (3 ppm) and the Knight shift of the TiDz phase
(−143 ppm). Exchange NMR indicates complete deuterium exchange at 25 °C between the MgD2 and TiDy phase within 1 s, as consistent with intimate contacts between these phases.
Combined analysis of the XRD and NMR peak areas suggests that the deuterium concentrations
y and z in the TiDyand TiDzdomains are about 1.5 and 2.0, respectively. Comparing the intrinsic
cell parameters of rutile MgH2 and fluorite TiH2, we propose that stabilization of the mixed
nanocomposite may arise from a coherent coupling between the crystal structures of the rutile MgD2 nanodomains and the thin layers of fcc TiDy.
*
This chapter is reproduced from S. Srinivasan, P.C.M.M. Magusin, W.P. Kalisvaart, P.H.L. Notten, F. Cuevas, M. Latroche, R.A. van Santen, , Phys. Rev. B: Condens. Matter Mater. Phys., 81(5), 054107-1, (2010).
3.1 Introduction
The increasing demand for energy puts the conventional fossil fuel sources under stress as these are already running out. An attractive alternative energy carrier would be hydrogen, if it can be efficiently stored. This can be achieved by means of pressurized gas or liquefied hydrogen. However, high pressure (typically 70 Mpa) or low temperature (21 K) are necessary to reach practical capacities. Alternative options at lower pressure and higher temperature are physical hydrogen storage in porous materials1,2 or chemical hydrogen storage in the form of hydride compounds, such as reversible metal hydrides.3 Metal-based hydrogen storage materials have the advantage of being operable under closer to ambient conditions when compared to other types of storage. The typical usage would be for stationary hydrogen storage or for electrode materials in rechargeable batteries for portable devices such as cell phones and cameras. Since hydrogen is absorbed in metal hydrides as atoms, rather than molecules, these hydrides typically have a volumetric hydrogen density similar to or higher than liquid hydrogen (71 kg m−3 at 20 K).4 Magnesium hydride, for instance, can theoretically store 110 kg m−3, and because magnesium is a light metal, it also has a high gravimetric storage capacity of 7.6 wt. % hydrogen. However, MgH2 suffers from slow sorption kinetics.
Notten and co-workers5–7 have recently found that the kinetics can be improved by alloying magnesium with scandium. A Sc fraction > 20 at. % transition metal causes a change from the rutile structure of MgH2 to a fluorite structure of transition-metal hydrides and thereby
enhances the mobility of the hydrogen.8 As revealed by NMR relaxometry,9 hydrogen mobility is significantly faster in Mg0.65Sc0.35H2.2 than in MgH2 and ScH2. Since scandium is a precious
metal, an economically more feasible option is to use its less expensive neighbor in the periodic table of elements, titanium, which has similar electrochemical properties in Mg-based thin films as scandium.10 Magnesium and titanium are thermodynamically immiscible,11 but alloys can be produced, e.g., in the form of thin films by e-beam deposition,10 physical vapor deposition12 and magnetron cosputtering,13 or as bulk powders by mechanical mixing.14–19 The boiling point of Mg is less than the melting point of Ti which rules out melt mixing techniques. The crystal structure of pure Mg and Ti is hexagonal, but ball milling of mixed Mg and Ti powders with micron-sized particles yields a material with two face-centered cubic (fcc) crystal phases.18 This is promising for its properties as a hydrogen storage material, because an fcc Mg-Ti lattice with hydrogen at the tetrahedral (T) interstitial sites would form the favorable fluorite ternary
